## Weather Shocks and Exchange Rate Flexibility — Executive Summary (wpiea2022093-print-pdf)

## Source details

**Canonical URL:** [Weather Shocks and Exchange Rate Flexibility — Executive Summary (wpiea2022093-print-pdf)](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022093-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022093-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022093-print-pdf.pdf.json)

---

### Research question and approach
- Key question: Does greater exchange rate flexibility help an economy’s adjustment to weather shocks?
- Weather shocks defined as annual variations in temperature and precipitation (based on Acevedo and others (2020)); treated as exogenous to exchange rate policy.
- Empirical strategy:
  - Local projections (Jordà, 2005) to trace impulse responses of real per capita GDP to weather shocks over horizons h = 0 to 5 years.
  - Regressions include one lag of dependent and independent variables, forwards of shock variables, country fixed effects, and time fixed effects interacted with region dummies.
  - Main coefficient of interest: β3h on the interaction between the shock and the regime indicator; under a peg Regimect = 1, effect under peg is (β2h + β3h), effect under non-peg is β2h.

### Key quantitative findings
- Exchange rate regime differences (temperature shocks):
  - A 1 degree Celsius increase in temperature decreases real per capita income growth by:
    - –1.395 percentage points upon impact under a pegged exchange rate regime (Horizon 0, Pegged).
    - –1.034 percentage points upon impact under a non-pegged exchange rate regime (Horizon 0, Nonpeg).
  - Medium-term (average over the five-year horizon) adverse growth impact:
    - Pegged: –1.4 percentage points (average over five-year horizon reported as about –1.4).
    - Flexible: –0.6 percentage points (less than one half that amount).
- Horizon-by-horizon per capita income growth (EMDC; Table 2 highlights):
  - Horizon 0: Nonpeg –1.034 ***, Pegged –1.395 ***, p_value 0.109
  - Horizon 1: Nonpeg –0.589, Pegged –1.219 **, p_value 0.050
  - Horizon 2: Nonpeg –0.477, Pegged –1.156 **, p_value 0.109
  - Horizon 3: Nonpeg –0.662, Pegged –1.343 **, p_value 0.203
  - Horizon 4: Nonpeg –0.852, Pegged –1.912 **, p_value 0.075
  - Horizon 5: Nonpeg –0.285, Pegged –1.572, p_value 0.048
- Export responses (medium-term):
  - Under flexible exchange rate regimes: real export growth increases by 1.4 percentage points.
  - Under pegs: real export growth declines by 5.5 percentage points.
- Historical/regime context:
  - Share of de facto hard pegs declined to 37 percent of the total number of countries in the early 1990s, and this share has increased to approximately 50 percent in recent years.
- Climatic background:
  - Global temperatures have increased by roughly 1 degree Celsius over the past century (IPCC, 2021).

### Mechanisms and channels of adjustment
- Exchange rate channel (Mundell‑Fleming type):
  - Under flexible regimes, nominal depreciation facilitates an export-driven recovery after adverse weather shocks.
  - Non-pegged regimes show significant real export growth responses consistent with depreciation.
- Inflation and real wages:
  - Inflation trends upward after temperature shocks under both regimes; inflation rises to a slightly greater extent under floating regimes with results at h=3 statistically significant.
  - Indicative real wage results: statistically significant two years after shocks under a peg relative to the float (data-limited; see Table A13).
- Fiscal channel and debt constraints:
  - Fiscal policy appears procyclical under both regimes; declines in government expenditure growth are statistically significant in first and third years after shocks when exchange rate is pegged (Table A14).
  - Larger number of high-debt countries implement fixed exchange rate regimes, suggesting constrained fiscal space limits countercyclical responses under pegs.
- Real Effective Exchange Rate (REER):
  - Both regimes experience real depreciations after temperature shocks; depreciation more pronounced under non-pegged regimes and statistically significant over the medium term (Table A12).

### Robustness, heterogeneity, and sample variations
- Sample and summary statistics (full sample highlights, Table 1):
  - Temperature (degree Celsius): N 13,170; Mean 18.9; SD 7.7; p50 21.4
  - Precipitation (100 millimeter per year): N 13,962; Mean 11.6; SD 8.4; p50 9.8
  - Per capita income growth: N 9,842; Mean 1.8; SD 6.5; p50 2.2
  - Government Debt (percent of GDP): N 7,560; Mean 55.3; SD 58.7; p50 42.3
- Sub-sample changes:
  - Temperature increased by roughly 1 degree Celsius between pre-1990 and post-1990 samples; result statistically significant at the 1 percent level.
  - Pre-1990 vs post-1990: post-1990 results resemble baseline for temperature shocks; precipitation shocks gain short- and medium-run significance post-1990.
- Robustness checks:
  - Full sample including AEs and EMDCs with average temperatures below 15 degrees Celsius: impact under peg for a 1-degree Celsius increase is 0.7 percentage points (full sample), roughly half the baseline EMDC-only estimate.
  - Constant-regime sample (regime identical year before, during, after shock) yields similar results to baseline.
  - Alternative exchange rate classification (Shambaugh (2004) updated): broadly similar results.
  - Using 3 lags instead of 1: coefficients under peg larger in absolute value and statistically significant at the one percent level in that specification (Table A8).
  - Including both shocks (temperature and precipitation) in the same regression produces similar coefficients and significance to single-shock specifications (Table 3).

### Policy implications and recommendations
- Exchange rate flexibility:
  - Greater exchange rate flexibility generally helps EMDCs better adjust to weather shocks by mitigating initial impact and promoting faster recovery via real exchange rate adjustment and higher real export growth.
  - Fixed exchange rate regimes can still be appropriate in specific circumstances (to "import" monetary credibility, reduce transaction costs, or where liability dollarization makes pegs welfare enhancing).
  - Adoption of greater flexibility should ensure a smooth transition and establishment of a new, credible nominal anchor.
- Fiscal policy and buffers:
  - Pegged regimes are expected to use countercyclical fiscal policy, but high public debt can restrain this response; building precautionary fiscal buffers is important.
  - Exchange rate regime choice must be considered alongside the policy mix and country-specific circumstances.
- Distributional and country-specific considerations:
  - Hot countries (overwhelmingly EMDCs) suffer more from temperature increases; agricultural dependence amplifies adverse effects under pegs.
  - Policymakers should weigh the potential sizable costs of pegs in EMDCs that frequently experience large adverse shocks if weather shocks resemble other shocks.

### Data, methodology, and limitations
- Data sources:
  - Temperature and precipitation: Acevedo and others (2020); CRU grid-cell data aggregated to country level using 1950 population weights.
  - Macroeconomic series: IMF World Economic Outlook (WEO) and World Bank World Development Indicators (WDI); exchange rate regimes: Ilzetzki, Reinhart, and Rogoff (2019), Shambaugh (2004), AREAER.
  - Real wages: International Labor Organization, Real wage database.
- Identification strengths:
  - Weather shocks treated as plausibly exogenous; country fixed effects and region-time fixed effects used to control for confounders.
- Caveats:
  - Results reflect averages and heterogeneity across countries exists.
  - Estimation intentionally parsimonious to avoid overcontrolling for variables potentially affected by weather shocks.

*Source: IMF Working Paper — Weather Shocks and Exchange Rate Flexibility (wpiea2022093-print-pdf).*

### Executive Summary

### Executive Summary

### Research question and approach
- Key question: Does greater exchange rate flexibility help an economy’s adjustment to weather shocks?
- Weather shocks: annual variations in temperature and precipitation (based on Acevedo and others (2020)); treated as exogenous to exchange rate policy.
- Empirical strategy: local projections (Jordà, 2005) to trace impulse responses of real per capita GDP to weather shocks over horizons h = 0 to 5 years; regressions include one lag of dependent and independent variables, forwards of shock variables, country fixed effects, and time fixed effects interacted with region dummies.
- Main coefficient of interest: β3h on the interaction between the shock and the regime indicator; under a peg Regimect = 1, effect under peg is (β2h + β3h), effect under non-peg is β2h.

### Key quantitative findings
- Exchange rate regime differences:
  - A 1 degree Celsius increase in temperature would decrease real per capita income growth by 1.4 percentage points upon impact under a peg.
  - The corresponding decrease under a non-pegged exchange rate regime would be 1.0 percentage point upon impact.
  - Medium-term adverse growth impact under the peg is about –1.4 percentage points on average.
  - Medium-term impact under flexible regimes is –0.6 percentage points on average (reported as less than one half that amount) and is not statistically significant.
- Export response (medium-term):
  - Under flexible exchange rate regimes: real export growth increases by 1.4 percentage points.
  - Under pegs: real export growth declines by 5.5 percentage points.
- Historical/regime context:
  - Share of de facto hard pegs declined to 37 percent of the total number of countries in the early 1990s, and this share has increased to approximately 50 percent in recent years.
- Geographic and climatic background:
  - Global temperatures have increased by roughly 1 degree Celsius over the past century (IPCC, 2021).

### Adjustment mechanisms and channels
- Evidence consistent with textbook Mundell-Fleming-type adjustment:
  - Under flexible regimes, depreciation of the nominal exchange rate facilitates an export-driven recovery following adverse weather shocks.
  - Non-pegged regimes show significant export responses consistent with this depreciation channel.
- Fiscal channel:
  - Under pegged regimes, a countercyclical fiscal policy stance would be expected in response to weather shocks, but high debt prevalence appears to restrain the use and effectiveness of fiscal stabilization.
  - This highlights the importance of the policy mix and precautionary (fiscal) buffers.

### Policy implications and heterogeneity
- Greater exchange rate flexibility generally helps EMDCs better adjust to weather shocks by mitigating initial impact and promoting faster recovery.
- Fixed exchange rate regimes can still be appropriate for some countries under certain circumstances (e.g., to "import" monetary credibility, reduce transaction costs, or when liability dollarization makes pegs welfare enhancing).
- Adoption of greater exchange rate flexibility should ensure a smooth transition and a new, credible nominal anchor.
- The costs of pegs could be sizeable in EMDCs that frequently experience large adverse shocks if weather shocks resemble other shocks.

### Methodology, data, and robustness considerations
- Data sources: IMF World Economic Outlook (WEO), World Bank World Development Indicators (WDI), University of East Anglia’s Climate Research Unit (CRU) historical temperature and precipitation databases; temperature/precipitation series constructed by aggregating CRU grid-cell data to country level using 1950 population weights (Acevedo and others (2020)).
- Identification strengths:
  - Weather shocks treated as plausibly exogenous and not directly caused by exchange rate choices.
  - Country fixed effects control for time-invariant country characteristics (latitude, size, average growth); region-time fixed effects account for region-specific annual shocks.
- Advantages over earlier studies:
  - Weather shocks are continuous measures (temperature, precipitation) that capture intensity, avoiding binary limitations of natural-disaster indicators.
  - Extended set of variables and inclusion of country and region-time fixed effects; use of local projections to trace medium-term dynamics.
- Caveats:
  - Role of heterogeneity across countries; results reflect averages and may not apply uniformly.
  - Estimation intentionally parsimonious to avoid overcontrolling for variables that may themselves be affected by weather shocks.

*Source: IMF WORKING PAPERS — Weather Shocks and Exchange Rate Flexibility: Executive Summary.*

### 0.7 degree Celsius higher in emerging market and developing countries (EMDCs). However, although most of

### wpiea2022093-print-pdf

### Main findings
- The implications on growth associated with adverse weather shocks are worse under fixed exchange rate regimes, especially for temperature shocks.
- Dynamic responses under pegged regimes are statistically significant upon impact and over the following three years; differences between pegged and non-pegged regimes are statistically significant in the year following the shock (gray shading in figures denotes significance at least at 10 percent).
- For temperature shocks, greater exchange rate flexibility both mitigates the initial impact and promotes a faster recovery.

### Key quantitative estimates (temperature shocks)
- A 1-degree Celsius increase in temperature decreases real per capita income growth by:
  - –1.395 percentage points upon impact under a pegged exchange rate regime (Horizon 0, Pegged).
  - –1.034 percentage points upon impact under a non-pegged exchange rate regime (Horizon 0, Nonpeg).
- Medium-term (average over the five-year horizon) adverse growth impact:
  - Pegged: –1.4 percentage points (average over five-year horizon reported as about –1.4).
  - Flexible: –0.6 percentage points (less than one half that amount).
- Table 2 horizon-by-horizon estimates (Per Capita Income Growth: EMDC):
  - Horizon 0: Nonpeg –1.034 ***, Pegged –1.395 ***, p_value 0.109
  - Horizon 1: Nonpeg –0.589, Pegged –1.219 **, p_value 0.050
  - Horizon 2: Nonpeg –0.477, Pegged –1.156 **, p_value 0.109
  - Horizon 3: Nonpeg –0.662, Pegged –1.343 **, p_value 0.203
  - Horizon 4: Nonpeg –0.852, Pegged –1.912 **, p_value 0.075
  - Horizon 5: Nonpeg –0.285, Pegged –1.572, p_value 0.048
- Table 3 (both shocks in same regression) highlights similar estimates:
  - Temp: Nonpeg Horizon 0 –1.095 ***, Horizon 1 –0.693, Horizon 2 –0.587, Horizon 3 –0.828, Horizon 4 –1.025, Horizon 5 –0.400
  - Temp: Peg Horizon 0 –1.406 ***, Horizon 1 –1.204 ***, Horizon 2 –1.144 **, Horizon 3 –1.313 **, Horizon 4 –1.856 **, Horizon 5 –1.500
  - Precip: Nonpeg coefficients small positive; Precip: Peg coefficients small negative (see Table 3 for standard errors and significance).
- Sample coverage for baseline: all Emerging Market and Developing Economies (EMDCs) with annual average temperature greater than 15 degrees Celsius; Number of countries 110; Number of observations 3268; Adjusted R2 reported (e.g., 0.34 for Table 2).

### Summary statistics (Table 1)
- Sample sizes and moments (N, Mean, SD, percentiles) for full sample:
  - Temperature (degree Celsius): N 13,170; Mean 18.9; SD 7.7; p1 1.2; p5 5.6; p25 11.9; p50 21.4; p75 25.5; p95 27.7; p99 28.8
  - Precipitation (100 millimeter per year): N 13,962; Mean 11.6; SD 8.4; p1 0.0; p5 0.0; p25 5.4; p50 9.8; p75 16.4; p95 28.1; p99 34.8
  - Per capita income growth: N 9,842; Mean 1.8; SD 6.5; p1 –17.2; p5 –7.2; p25 –0.3; p50 2.2; p75 4.5; p95 9.3; p99 17.0
  - Inflation: N 7,368; Mean 25.8; SD 351.4; p1 –3.0; p5 –0.3; p25 2.3; p50 5.2; p75 10.8; p95 38.6; p99 235.6
  - Exports (percent of GDP): N 7,882; Mean 36.0; SD 28.3; p1 4.2; p5 8.1; p25 18.1; p50 28.9; p75 45.8; p95 82.5; p99 158.4
  - Real exports growth: N 5,722; Mean –3.7; SD 31.6; p1 –120.1; p5 –41.6; p25 –10.1; p50 0.9; p75 9.8; p95 24.2; p99 44.5
  - Real Effective Exchange Rate growth: N 5,919; Mean –0.4; SD 13.0; p1 –39.8; p5 –14.3; p25 –3.4; p50 0.3; p75 3.7; p95 12.0; p99 26.0
  - Real Wage Growth: N 1,938; Mean 3.1; SD 7.5; p1 –14.0; p5 –5.5; p25 –0.3; p50 1.7; p75 5.0; p95 16.4; p99 27.5
  - Real Government Consumption growth: N 7,077; Mean 4.0; SD 13.6; p1 –37.3; p5 –14.9; p25 –0.3; p50 3.9; p75 8.5; p95 22.0; p99 45.1
  - Government Debt (percent of GDP): N 7,560; Mean 55.3; SD 58.7; p1 2.7; p5 7.8; p25 23.7; p50 42.3; p75 68.8; p95 141.2; p99 260.5
- Sub-sample split pre-1990 (1950-1989) vs post-1990 (1990-2015) shows:
  - Temperature increased by roughly 1 degree Celsius between the two sub-samples; result statistically significant at the 1 percent level.
  - Pre 1990 Temperature: N 7,980; Mean 18.6; SD 7.7; p50 21.1.
  - Post 1990 Temperature: N 5,190; Mean 19.4; SD 7.7; p50 21.9.
  - Per capita income growth: Pre 1990 Mean 1.9; Post 1990 Mean 1.8.

### Robustness checks and sample variations
- Broader full sample including Advanced Economies and EMDCs with average temperatures below 15 degrees Celsius:
  - Impact under peg for a 1-degree Celsius increase is 0.7 percentage points (full sample), roughly half the baseline EMDC-only estimate.
- Constant-regime robustness: restricting to shocks where the exchange rate regime remained identical over a three-year window (year before the shock, shock year, year after) yields results similar to baseline.
- Alternative exchange rate classification: Shambaugh (2004) updated dataset yields broadly similar results (negative coefficients under peg, statistically significant upon impact and over medium term).
- Lag dynamics: using 3 lags instead of 1 in regressions yields robust results; coefficients under peg are larger in absolute value and all coefficients are statistically significant at the one percent level in that specification.
- Both shocks included together (temperature and precipitation) produce remarkably similar coefficients and statistical significance to single-shock specifications (Table 3).

### Mechanisms identified
- Inflation:
  - Inflation trends upward after temperature shocks under both regimes; inflation rises to a slightly greater extent under floating regimes; results three years after the shock (h=3) are statistically significant.
- Exports:
  - Export-to-GDP ratio shows broadly similar responses under both regimes, but this may reflect GDP behavior (denominator effects).
  - Real export growth shows clear regime differences:
    - Under pegged regimes, real export growth displays a downward trend after temperature shocks.
    - Under non-pegged regimes, a J-curve-like pattern appears with increases in real export growth.
    - One year after the shock: flexible regimes see +0.8 percentage points in real export growth; pegged regimes see –3.4 percentage points.
    - Medium-term: flexible regimes show +1.4 percentage points; pegs show –5.5 percentage points (decline).
- Real Effective Exchange Rate (REER):
  - Both regimes show real depreciations after temperature shocks; depreciation is more pronounced under non-pegged regimes and statistically significant over the medium term.
- Real wages:
  - Indicative results (data-limited) suggest real wages are statistically significant two years after shocks under a peg relative to the float (see Appendix Table A13).

### Fiscal policy, debt, and policy mix
- Fiscal response measured by real growth in government expenditures:
  - Fiscal policy appears procyclical under both regimes; declines in government expenditure growth are statistically significant in first and third years after shocks when exchange rate is pegged.
  - Under flexible regimes the decline is not statistically significant (could be interpreted as acyclical fiscal stance).
- Debt constraints:
  - There is a larger number of high-debt countries that implement fixed exchange rate regimes (counts above sample mean, median, and 75th percentile shown in figure), suggesting constrained fiscal space may limit countercyclical fiscal responses under pegs.
- Policy implication: the choice and effectiveness of the exchange rate regime must be considered alongside the policy mix and precautionary fiscal buffers.

### How results vary over time and by country groups
- Pre-1990 sample:
  - Both temperature and precipitation shocks have similar growth implications; only temperature shocks have a short-term statistically significant impact.
- Post-1990 sample:
  - Results resemble the full baseline: temperature shocks show short- and medium-term negative impacts similar to baseline.
  - Precipitation shocks become statistically significant in the short-run (upon impact) and over the medium term (e.g., h=4) in the post-1990 sample.
  - Differences in growth outcomes between fixed and flexible regimes for precipitation shocks become statistically significant post-1990.
- Heterogeneity:
  - Hot countries (overwhelmingly EMDCs) suffer more from temperature increases; baseline sample restriction to countries with annual average temperature greater than 15 degrees Celsius is motivated by this heterogeneity.
  - For countries where agricultural sector as share of GDP is above the sample median, impact under the peg can be more severe (medium-term impact coefficients larger in absolute value and statistically significant at the 5 percent level; results available upon request in source).

### Policy recommendations and concluding remarks
- Greater exchange rate flexibility can help EMDCs better adjust to weather shocks by:
  - Mitigating the initial adverse impact of temperature shocks on per capita income growth.
  - Promoting a faster recovery through real exchange rate adjustment, higher real export growth, and inflation pass-through.
- Fiscal policy matters:
  - Pegged regimes are expected to rely on countercyclical fiscal policy, but high debt can limit this response; building precautionary fiscal buffers is important.
- Policy design considerations:
  - The adoption of greater exchange rate flexibility should ensure a smooth transition and be supported by a new, credible nominal anchor.
  - Exchange rate regime choice cannot be considered in isolation from other macroeconomic policies and country-specific circumstances; fixed regimes may still have advantages for some countries under certain conditions.

*Source: IMF Working Paper — Weather Shocks and Exchange Rate Flexibility (authors’ calculations and tables/figures as presented in the supplied content).*

### Appendix A

### Appendix A — Data Sources and Country Groupings

### Data sources (Table A1)
- Temperature and precipitation: Acevedo and others (2020)
- Per capita income: IMF, World Economic Outlook database; World Bank, World Development Indicators database
- Exchange rate regimes: Ilzetzki, Reinhart, and Rogoff (2019), Shambaugh (2004) Exchange Rate Regime Classification, IMF Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER)
- Real Exports: IMF, World Economic Outlook database; World Bank, World Development Indicators database
- Government Expenditure: IMF, World Economic Outlook database; World Bank, World Development Indicators database
- Inflation: IMF, World Economic Outlook database; World Bank, World Development Indicators database
- Government Debt: IMF, World Economic Outlook database; World Bank, World Development Indicators database
- Real Effective Exchange Rate: IMF, World Economic Outlook database; World Bank, World Development Indicators database
- Real Wages: International Labor Organization, Real wage database

### Country groupings (Table A2)
- Advanced Economies: Australia, Austria, Belgium, Canada, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Iceland, Ireland, Israel, Italy, Japan, Korea, Latvia, Lithuania, Luxembourg, Malta, Netherlands, New Zealand, Norway, Portugal, Puerto Rico, Singapore, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, United Kingdom, United States
- Emerging Market Economies: Albania, Algeria, Angola, Antigua and Barbuda, Argentina, Armenia, Azerbaijan, Bahrain, Barbados, Belarus, Belize, Bosnia and Herzegovina, Botswana, Brazil, Brunei Darussalam, Bulgaria, Cabo Verde, Chile, China, Colombia, Costa Rica, Croatia, Dominica, Dominican Republic, Ecuador, Egypt, El Salvador, Equatorial Guinea, Fiji, Gabon, Georgia, Grenada, Guatemala, Guyana, Hungary, India, Indonesia, Iran, Iraq, Jamaica, Jordan, Kazakhstan, Kuwait, Lebanon, Libya, Malaysia, Mauritius, Mexico, Montenegro, Morocco, Namibia, North Macedonia, Oman, Pakistan, Panama, Paraguay, Peru, Philippines, Poland, Qatar, Romania, Russia, Samoa, Saudi Arabia, Serbia, South Africa, Sri Lanka, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, Suriname, Swaziland, Syria, Thailand, Timor-Leste, Tonga, Trinidad and Tobago, Tunisia, Turkey, Turkmenistan, Ukraine, United Arab Emirates, Uruguay, Vanuatu, Venezuela
- Developing Countries: Afghanistan, Bangladesh, Benin, Bhutan, Bolivia, Burkina Faso, Burundi, Cambodia, Cameroon, Central African Republic, Chad, Comoros, Democratic Republic of the Congo, Republic of Congo, Côte d'Ivoire, Djibouti, Eritrea, Ethiopia, The Gambia, Ghana, Guinea, Guinea-Bissau, Haiti, Honduras, Kenya, Kyrgyz Republic, Lao P.D.R., Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Moldova, Mongolia, Mozambique, Myanmar, Nepal, Nicaragua, Niger, Nigeria, Papua New Guinea, Rwanda, Senegal, Sierra Leone, Solomon Islands, South Sudan, Sudan, São Tomé and Príncipe, Tajikistan, Tanzania, Togo, Uganda, Uzbekistan, Vietnam, Yemen, Zambia, Zimbabwe

### Empirical results: Impact of temperature and precipitation shocks (Tables A3–A14)
- Table A3: The Impact of Temperature Shocks Over Time
  - Per Capita Income Growth: After 1990 (sample: annual average temperature > 15 0 C)
    - Horizon 0: Nonpeg -0.985 ** (0.387); Pegged -1.479 *** (0.462); p_value 0.124
    - Horizon 1: Nonpeg -0.508 (0.460); Pegged -1.273 ** (0.537); p_value 0.094
    - Horizon 2: Nonpeg -0.385 (0.546); Pegged -1.064 * (0.594); p_value 0.254
    - Horizon 3: Nonpeg -0.792 (0.695); Pegged -1.341 * (0.752); p_value 0.423
    - Horizon 4: Nonpeg -0.813 (0.884); Pegged -1.566 * (0.898); p_value 0.306
    - Horizon 5: Nonpeg -0.111 (0.914); Pegged -1.007 (0.925); p_value 0.250
    - Adjusted R2 0.41; Number of countries 115; Number of observations 2053
  - Per Capita Income Growth: Before 1990
    - Horizon 0: Nonpeg -1.427 * (0.744); Pegged -1.424 * (0.771); p_value 0.991
    - Horizon 1: Nonpeg -0.747 (1.139); Pegged -0.813 (1.118); p_value 0.838
    - Horizon 2: Nonpeg -0.585 (1.189); Pegged -0.840 (1.196); p_value 0.463
    - Horizon 3: Nonpeg -0.265 (1.473); Pegged -0.479 (1.496); p_value 0.585
    - Horizon 4: Nonpeg -1.279 (1.498); Pegged -1.638 (1.495); p_value 0.456
    - Horizon 5: Nonpeg -0.486 (1.519); Pegged -0.936 (1.480); p_value 0.368
    - Adjusted R2 0.49; Number of countries 78; Number of observations 1428

- Table A4: The Impact of Precipitation Shocks Over Time
  - Per Capita Income Growth: Before 1990
    - Horizon 0: Nonpeg 0.201 (0.137); Pegged 0.129 (0.092); p_value 0.514
    - Horizon 1: Nonpeg 0.090 (0.143); Pegged 0.045 (0.109); p_value 0.717
    - Horizon 2: Nonpeg 0.002 (0.179); Pegged -0.036 (0.150); p_value 0.808
    - Horizon 3: Nonpeg 0.328 (0.214); Pegged 0.120 (0.183); p_value 0.216
    - Horizon 4: Nonpeg 0.499 * (0.273); Pegged 0.274 (0.226); p_value 0.216
    - Horizon 5: Nonpeg 0.442 * (0.261); Pegged 0.356 (0.251); p_value 0.614
    - Adjusted R2 0.49; Number of countries 78; Number of observations 1428
  - Per Capita Income Growth: After 1990
    - Horizon 0: Nonpeg 0.085 (0.096); Pegged -0.162 * (0.096); p_value 0.143
    - Horizon 1: Nonpeg 0.174 (0.146); Pegged -0.266 (0.175); p_value 0.074
    - Horizon 2: Nonpeg 0.234 (0.157); Pegged -0.218 (0.188); p_value 0.074
    - Horizon 3: Nonpeg 0.236 (0.162); Pegged -0.321 (0.202); p_value 0.046
    - Horizon 4: Nonpeg 0.222 (0.140); Pegged -0.400 ** (0.186); p_value 0.018
    - Horizon 5: Nonpeg 0.262 (0.167); Pegged -0.231 (0.240); p_value 0.069
    - Adjusted R2 0.41; Number of countries 115; Number of observations 2053

- Table A5: The Impact of Temperature Shocks on per Capita Income Growth for full set of Countries (Full Sample: Emerging Market and Developing Economies)
  - Horizon 0: Nonpeg -0.506 *** (0.184); Pegged -0.668 *** (0.200); p_value 0.043
  - Horizon 1: Nonpeg -0.405 * (0.229); Pegged -0.630 ** (0.259); p_value 0.069
  - Horizon 2: Nonpeg -0.308 (0.293); Pegged -0.586 * (0.331); p_value 0.054
  - Horizon 3: Nonpeg -0.393 (0.351); Pegged -0.658 (0.413); p_value 0.134
  - Horizon 4: Nonpeg -0.336 (0.449); Pegged -0.569 (0.498); p_value 0.223
  - Horizon 5: Nonpeg -0.238 (0.504); Pegged -0.512 (0.555); p_value 0.186
  - Adjusted R2 0.39; Number of countries 173; Number of observations 5182

- Table A6: The Impact of Temperature Shocks on per Capita Income Growth for “Constant Regime” (EMDC; sample: EMDCs with annual average temperature > 15 0 C)
  - Horizon 0: Nonpeg -0.996 ** (0.456); Pegged -1.416 *** (0.433); p_value 0.448
  - Horizon 1: Nonpeg -0.215 (0.668); Pegged -1.600 *** (0.537); p_value 0.059
  - Horizon 2: Nonpeg 0.008 (0.727); Pegged -1.623 ** (0.740); p_value 0.097
  - Horizon 3: Nonpeg -0.339 (0.728); Pegged -1.602 ** (0.756); p_value 0.109
  - Horizon 4: Nonpeg -0.469 (0.935); Pegged -2.273 ** (1.107); p_value 0.120
  - Horizon 5: Nonpeg 0.224 (1.007); Pegged -2.090 * (1.156); p_value 0.071
  - Adjusted R2 0.35; Number of countries 110; Number of observations 3153

- Table A7: The Impact of Temperature Shocks on per Capita Income Growth using Shambaugh’s Classification (EMDC; sample: EMDCs with annual average temperature > 15 0 C)
  - Horizon 0: Nonpeg -0.824 *** (0.310); Pegged -1.089 *** (0.303); p_value 0.179
  - Horizon 1: Nonpeg -0.647 (0.433); Pegged -0.958 ** (0.394); p_value 0.166
  - Horizon 2: Nonpeg -0.767 (0.463); Pegged -1.043 ** (0.437); p_value 0.251
  - Horizon 3: Nonpeg -1.068 ** (0.429); Pegged -1.329 *** (0.483); p_value 0.398
  - Horizon 4: Nonpeg -1.299 ** (0.599); Pegged -1.848 *** (0.644); p_value 0.193
  - Horizon 5: Nonpeg -1.082 (0.743); Pegged -1.442 ** (0.701); p_value 0.255
  - Adjusted R2 0.30; Number of countries 108; Number of observations 4950

- Table A8: The Impact of Temperature Shocks on per Capita Income Growth (3 lags; EMDC; sample: EMDCs with annual average temperature > 15 0 C)
  - Horizon 0: Nonpeg -1.141 *** (0.352); Pegged -1.546 *** (0.344); p_value 0.085
  - Horizon 1: Nonpeg -0.631 (0.513); Pegged -1.330 *** (0.478); p_value 0.063
  - Horizon 2: Nonpeg -0.440 (0.499); Pegged -1.330 ** (0.536); p_value 0.076
  - Horizon 3: Nonpeg -0.763 (0.568); Pegged -1.703 ** (0.651); p_value 0.127
  - Horizon 4: Nonpeg -1.151 * (0.671); Pegged -2.284 *** (0.728); p_value 0.079
  - Horizon 5: Nonpeg -0.941 (0.656); Pegged -2.184 *** (0.766); p_value 0.060
  - Adjusted R2 0.35; Number of countries 115; Number of observations 3252

- Table A9: The Impact of Temperature Shocks on Inflation (Inflation: EMDC; sample: EMDCs with annual average temperature > 15 0 C)
  - Horizon 0: Nonpeg 5.264 (5.264); Pegged 4.220 (5.203); p_value 0.616
  - Horizon 1: Nonpeg 6.162 (4.770); Pegged 3.661 (5.128); p_value 0.197
  - Horizon 2: Nonpeg 6.837 (4.956); Pegged 5.057 (5.416); p_value 0.515
  - Horizon 3: Nonpeg 9.986 ** (4.926); Pegged 9.145 (5.589); p_value 0.747
  - Horizon 4: Nonpeg 8.429 (5.988); Pegged 5.530 (6.283); p_value 0.273
  - Horizon 5: Nonpeg 12.031 * (6.194); Pegged 12.047 * (7.069); p_value 0.996
  - Adjusted R2 0.29; Number of countries 103; Number of observations 2316

- Table A10: The Impact of Temperature Shocks on Export-to-GDP Ratio (Export-to-GDP Ratio: EMDC; sample: EMDCs with annual average temperature > 15 0 C)
  - Horizon 0: Nonpeg 1.054 (1.688); Pegged 0.478 (1.878); p_value 0.398
  - Horizon 1: Nonpeg 4.673 ** (2.034); Pegged 4.186 ** (1.895); p_value 0.470
  - Horizon 2: Nonpeg 2.817 (2.592); Pegged 2.569 (2.322); p_value 0.775
  - Horizon 3: Nonpeg 2.948 (2.911); Pegged 2.857 (2.978); p_value 0.928
  - Horizon 4: Nonpeg 2.804 (2.598); Pegged 2.794 (2.939); p_value 0.994
  - Horizon 5: Nonpeg 2.823 (2.131); Pegged 1.705 (1.903); p_value 0.334
  - Adjusted R2 0.11; Number of countries 101; Number of observations 2887

- Table A11: The Impact of Temperature Shocks on Real Export Growth (Real Export Growth: EMDC; sample: EMDCs with annual average temperature > 15 0 C)
  - Horizon 0: Nonpeg -1.774 (1.465); Pegged -2.714 (1.989); p_value 0.367
  - Horizon 1: Nonpeg 0.844 (2.474); Pegged -3.404 (3.581); p_value 0.057
  - Horizon 2: Nonpeg 2.119 (3.355); Pegged -4.044 (4.297); p_value 0.108
  - Horizon 3: Nonpeg 0.483 (4.100); Pegged -8.444 (6.495); p_value 0.140
  - Horizon 4: Nonpeg 2.906 (4.251); Pegged -7.985 (7.377); p_value 0.099
  - Horizon 5: Nonpeg 4.005 (4.288); Pegged -6.413 (7.414); p_value 0.121
  - Adjusted R2 0.49; Number of countries 104; Number of observations 2282

- Table A12: The Impact of Temperature Shocks on Real Effective Exchange Rate (REER: EMDC; sample: EMDCs with annual average temperature > 15 0 C)
  - Horizon 0: Nonpeg -0.447 (0.959); Pegged 0.086 (0.855); p_value 0.487
  - Horizon 1: Nonpeg -1.230 (1.694); Pegged -1.292 (1.602); p_value 0.947
  - Horizon 2: Nonpeg -1.605 (1.561); Pegged -1.182 (1.547); p_value 0.745
  - Horizon 3: Nonpeg -1.918 (1.682); Pegged -1.817 (1.561); p_value 0.939
  - Horizon 4: Nonpeg -2.724 * (1.637); Pegged -2.434 (1.639); p_value 0.828
  - Horizon 5: Nonpeg -1.479 (1.835); Pegged -0.826 (1.759); p_value 0.632
  - Adjusted R2 0.40; Number of countries 114; Number of observations 2534

- Table A13: The Impact of Temperature Shocks on Real Wages (Real Wages: EMDC; sample: EMDCs with annual average temperature > 15 0 C)
  - Horizon 0: Nonpeg 0.465 (0.620); Pegged 0.057 (0.717); p_value 0.196
  - Horizon 1: Nonpeg -0.464 (0.473); Pegged -0.090 (0.611); p_value 0.299
  - Horizon 2: Nonpeg -0.105 (0.548); Pegged 0.592 (0.669); p_value 0.075
  - Horizon 3: Nonpeg 0.081 (0.424); Pegged 0.054 (0.411); p_value 0.884
  - Horizon 4: Nonpeg 0.264 (0.505); Pegged 0.182 (0.528); p_value 0.657
  - Horizon 5: Nonpeg 0.771 (0.530); Pegged 0.100 (0.572); p_value 0.047
  - Adjusted R2 0.66; Number of countries 79; Number of observations 701

- Table A14: The Impact of Temperature Shocks on Real Government Expenditure (Real Government Expenditure Growth: EMDC; sample: EMDCs with annual average temperature > 15 0 C)
  - Horizon 0: Nonpeg -0.703 (0.824); Pegged -1.189 (0.831); p_value 0.463
  - Horizon 1: Nonpeg -0.820 (1.317); Pegged -2.013 (1.263); p_value 0.129
  - Horizon 2: Nonpeg 0.339 (1.225); Pegged -1.465 (1.570); p_value 0.172
  - Horizon 3: Nonpeg -1.725 (1.449); Pegged -3.761 * (2.010); p_value 0.148
  - Horizon 4: Nonpeg -0.001 (1.454); Pegged -1.284 (1.703); p_value 0.336
  - Horizon 5: Nonpeg 1.860 (1.762); Pegged 0.827 (1.946); p_value 0.423
  - Adjusted R2 0.15; Number of countries 94; Number of observations 2638

*Source: Authors’ calculations.*

### References

### References

### Major themes represented
- Weather shocks and economic activity in low-income and developing countries
- Temperature effects on economic production and allocation of time
- Catastrophic natural disasters and macroeconomic outcomes
- Exchange rate regimes, flexibility, and adjustment to shocks
- Monetary policy, balance sheets, and capital controls in disaster-prone or open economies
- Time series and impulse response estimation methods
- Climate science assessment informing economic analysis

### Representative citations (verbatim examples from the source)
- Acevedo, S., Mrkaic, M., Novta, N., Poplawski-Ribeiro, M., Pugacheva, E., & Topalova, P. (2017). The Effects of weather shocks on economic activity: How can low-income countries cope. World Economic Outlook.
- Acevedo, S., Baccianti, C., Mrkaic, M., Novta, N., Pugacheva, E., & Topalova, P. (2019). Weather Shocks and Output in Low-Income Countries: The Role of Policies and Adaptation (No. 19/178). International Monetary Fund.
- Acevedo, S., Mrkaic, M., Novta, N., Pugacheva, E., & Topalova, P. (2020). The effects of weather shocks on economic activity: what are the channels of impact?. Journal of Macroeconomics, 65, 103207.
- Burke, M., Hsiang, S. M., & Miguel, E. (2015). Global non-linear effect of temperature on economic production. Nature, 527(7577), 235-239.
- Cavallo, E., Galiani, S., Noy, I., & Pantano, J. (2013). Catastrophic natural disasters and economic growth. Review of Economics and Statistics, 95(5), 1549-1561.
- IPCC, 2021: Summary for Policymakers. In: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Masson-Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Péan, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O. Yelekçi, R. Yu, and B. Zhou (eds.)]. In Press.
- Frankel, J., & Rose, A. (2002). An estimate of the effect of common currencies on trade and income. The quarterly journal of economics, 117(2), 437-466.
- Chinn, M. D., & Wei, S. J. (2013). A faith-based initiative meets the evidence: does a flexible exchange rate regime really facilitate current account adjustment?. Review of Economics and Statistics, 95(1), 168-184.
- Jordà, Ò. (2005). Estimation and inference of impulse responses by local projections. American economic review, 95(1), 161-182.
- Weitzman, M. L. (2011). Fat-tailed uncertainty in the economics of catastrophic climate change. Review of Environmental Economics and Policy, 5(2), 275-292.

### Methodological and policy-relevant sources highlighted
- Time series and forecasting methods: Box, G. E. P. and Jenkins, G. M. (1976). Time Series Analysis, Forecasting and Control.
- Impulse response estimation: Jordà, Ò. (2005). Estimation and inference of impulse responses by local projections.
- Exchange rate theory and regime choice: Mundell, R. A. (1961). A theory of optimum currency areas; Dornbusch, R. (1976). Expectations and exchange rate dynamics; Frankel, J. A. (2012). Choosing an exchange rate Regime.
- Macroeconomic policy in disaster contexts: Cantelmo, A., Fatouros, N., Melina, G., & Papageorgiou, C. (2022). Monetary Policy in Disaster-Prone Developing Countries.

*Weather Shocks and Exchange Rate Flexibility — Working Paper No. WP/2022/093*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022093-print-pdf.pdf_